Chest Radiography

Most pleural effusions can be recognized on posteroanterior (greater than 200 mL fluid volume) or lateral (greater than 50 mL fluid volume) chest radiographs. Large effusions may produce contralateral mediastinal shift. Blunting of the costophrenic angle may be the only sign for a small effusion but may also represent pleural thickening.

Thoracic Ultrasound Imaging

The availability of bedside thoracic ultrasound examination by clinicians has had a significant impact on pleural disease management in recent years. The 2010 British Thoracic Society Pleural Disease Guidelines strongly recommended the use of thoracic ultrasound imaging before procedures for pleural fluid. It is particularly useful for the detection (sensitivity approximately 100%), quantification (by depth), and characterization of pleural fluid (Figures 69-1 to 69-3; Table 69-2), as well as for guiding intervention. Ultrasonography is invaluable in the differentiation between pleural fluid and collapsed or consolidated lung, thereby avoiding unnecessary pleural procedures and associated complications.

Figure 69-1 Thoracic ultrasound image demonstrating a large anechoic effusion with compressive atelectasis of underlying lung.

Figure 69-2 A heavily septated effusion is evident on this ultrasound image.

Figure 69-3 Ultrasound appearance strongly suggestive of malignancy, with diaphragmatic nodularity and a large echogenic exudative pleural effusion.

Table 69-2 Pleural Fluid Sonographic Appearances

Thoracentesis guided by clinical examination alone could result in organ puncture in 10% of cases. Several large studies have demonstrated improved safety of pleural interventions performed under ultrasound guidance, particularly in reducing iatrogenic pneumothorax or organ puncture. A Mayo Clinic study showed a dramatic reduction (from 8% to 1%) in rate of thoracentesis-related complications since the unit initiated a “pleural safety program” that included pleural ultrasound training and mandated its use before thoracentesis. With adequate training in this modality, thoracic ultrasound imaging performed by respiratory physicians has been shown to have a safety profile comparable to that when performed by radiologists.

Ultrasound imaging has a high sensitivity (approximately 80%) for detecting pleural malignancy, which can manifest as thickening or nodularity on the visceral, parietal, and diaphragmatic pleural surfaces (see Figure 69-3). Detection of pleural nodularity mandates further investigation (e.g., with chest computed tomography [CT] and pleural biopsy), even if there are no further suspicious features. Ultrasonography also can identify abnormalities beyond the pleural cavity that may provide vital clues to the cause of the effusion, including peripheral lung tumors or abscesses, parenchymal consolidation and atelectasis, diaphragmatic paralysis or elevation, pericardial effusion, and rib and liver metastases and enables evaluation of supraclavicular and cervical lymphadenopathy (Figure 69-4).

Figure 69-4 Neck ultrasound image showing an enlarged, rounded, hypoechoic lymph node characteristic of malignancy, undergoing fine needle aspiration with a 21-gauge needle (*indicates needle tip).

Pleural interventions may be guided by site marking or real-time needle visualization; the latter is required to sample small or loculated effusions. The Royal College of Radiologists (in the United Kingdom), the American College of Chest Physicians, the American College of Surgeons, and the American College of Emergency Physicians are among the many agencies that have published ultrasound training guidelines for clinicians. Appropriate training is essential, because potential pitfalls with performance and interpretation of pleural ultrasound studies are recognized.

Computed Tomography and Magnetic Resonance Imaging

CT with pleural phase contrast enhancement highlights pleural abnormalities and aids discrimination of benign from malignant disease (see Chapter 7). Specific “pleural” CT protocols should be adopted for optimal pleural enhancement and abnormality detection; recent data suggest that images should be acquired 60 seconds after injection of 150 mL of an intravenous contrast agent at 2.5 mL/second. The presence of contraction of the hemithorax, mediastinal pleural involvement, and circumferential pleural thickening (especially greater than 1 cm and with nodularity) all are suggestive of pleural malignancy (Figure 69-5) but cannot adequately differentiate mesothelioma from metastatic pleural cancers. Magnetic resonance imaging (MRI) can help delineate malignant chest wall involvement and is valuable in selected cases, particularly when (probably benign) pleural abnormalities are to be followed clinically by serial imaging in younger patients.

Figure 69-5 Thoracic computed tomography (CT) scan showing circumferential nodular pleural thickening and hemithorax volume loss due to malignant pleural disease.

Positron Emission Tomography

Positron emission tomography (PET)-CT scanning (see Chapter 8) is beginning to emerge as a useful tool in pleural disease management. PET-CT cannot adequately differentiate between benign and malignant effusions, because of the tracer ¹⁸F-fluorodeoxyglucose (FDG). FDG-enhanced PET imaging is confounded by avid pleural uptake of FDG in the presence of pleural inflammation (including that due to previous talc pleurodesis and pleural infection). However, FDG-PET may have a role in guiding pleural biopsy in patients with diffuse pleural abnormality to increase sensitivity (Figure 69-6). FDG-PET also may identify nonpleural sites that allow tissue sampling to confirm malignancy (e.g., lymphadenopathy or liver metastases). Recent data suggest a role for FDG-PET in monitoring disease response to therapy in malignant mesothelioma, as well as a potential prognostic role.

Figure 69-6 Positron emission tomography–computed tomography (PET-CT) image demonstrating localized fluorodeoxyglucose (FDG)-avid mediastinal pleural thickening (arrow) in mesothelioma.

PET scanning using various novel molecular tracers is in early-phase trials for evaluation of pleural malignancies. For instance, PET scanning using labeled thymidine, essential for deoxyribonucleic acid (DNA) synthesis, can identify sites of cell proliferation activity and is not confounded by inflammation (Figure 69-7). New tracers targeting specific cell biology processes (e.g., annexin, a marker of apoptosis) are likely to provide valuable insight to disease pathobiology.

Figure 69-7 Positron emission tomography (PET) imaging using labeled thymidine in a patient with mesothelioma. Images before (A) and after (B) chemotherapy reveal dramatic improvement in the costal and mediastinal pleural uptake with treatment.

(Courtesy Dr. Roslyn Francis, University of Western Australia, Australia.)

Bedside Examination of the Fluid

Most pleural fluids are straw-colored (serous), serosanguineous (blood-stained), or hemorrhagic. A genuine hemothorax is characterized by fluid hematocrit greater than 50% of blood hematocrit.

Appearance of the fluid may reveal the diagnosis: presence of pus confirms empyema; food particles in the fluid suggest esophageal perforation; and chylous fluids suggest chylothorax or pseudochylothorax. Pleural fluid that appears milky should be centrifuged: in empyema, a clear supernatant is seen, whereas chyle remains cloudy. The odor of ammonia often is the most important diagnostic clue to the presence of urinothorax.

Separation of Exudates and Transudates

Exudative pleural effusions most commonly are defined by Light’s criteria (Box 69-1), using the fluid-to-serum ratio of protein and lactate dehydrogenase, which has an accuracy of 96%. Numerous other markers and criteria have been tested (including measurement of pleural fluid cholesterol values), but none has proved superior. Distinguishing exudates from transudates may narrow the scope of the differential diagnosis and streamline further investigations, although such categorization has limitations: It does not provide the diagnosis and fails to identify concurrent transudative and exudative causes of fluid formation. Research in recent years has focused on disease-specific markers that may provide a definitive diagnosis.

Differential Leukocyte Count

The cellular portion of physiologic pleural fluid consists predominantly of macrophages and monocytes. In disease states, the differential cell count of the pleural fluid may be helpful in determining the cause (Box 69-2). Acute pleural inflammation or injury generates chemotaxins, such as interleukin 8, and attracts neutrophils to the pleural space. A neutrophil-predominant effusion is commonly seen with acute bacterial pneumonia or pulmonary infarction. A lymphocyte-rich fluid is more common in disease of insidious onset such as tuberculosis (TB) or malignancy. Tuberculous effusions occasionally (less than 10%) may be neutrophilic. An increased eosinophil count (more than 10% of total leukocytes) is often nonspecific. Most commonly, eosinophil effusions develop secondary to presence of intrapleural air or blood (including pneumothorax or previous interventions) but can also be associated with a range of other diseases, such as Churg-Strauss syndrome or drug-induced pleuritis.

pH and Glucose

Pleural fluid pH (or glucose) measurement can aid disease management. Low glucose levels are associated with a similar spectrum of diseases that give rise to low pH effusions (e.g., infection and connective tissue diseases) (Box 69-3) and are equally informative except in patients with hyperglycemia.

Physiologic pleural fluid pH is approximately 7.6 and reflects bicarbonate accumulation within the pleural cavity. Pleural fluid pH should be measured using a blood gas analyzer, and care should be taken to avoid exposure of the fluid to free air or to residual lidocaine, which will potently raise or lower the fluid pH, respectively. pH meters and litmus paper have been shown to give unreliable values. Low pleural fluid pH (e.g., less than 7.3) coexistent with a normal blood pH may result from increased metabolism of leukocytes, bacteria, or tumor cells or hydrogen ion accumulation.

In malignant pleural effusions, a low pH is associated with more extensive tumor involvement of the pleura, a higher chance of positive findings on cytologic examination, a lower success rate for pleurodesis, and a poorer prognosis. In parapneumonic effusion, a low pH (less than 7.2) predicts the need for chest tube drainage, and pH below 7.2 is now often used as a cutoff point to diagnose pleural infection (complicated parapneumonic effusion) in the presence of a compatible clinical history. A recent study of 308 patients confirmed that pleural fluid pH (or glucose) measurements were superior to other biomarkers tested (pleural fluid procalcitonin, C-reactive protein, lipopolysaccharide-binding protein, and triggering receptor expressed on myeloid cells-1 [sTREM-1]) in distinguishing between simple and complicated parapneumonic effusions.

Disease-Specific Tests

The diagnosis of a malignant effusion should be established only by histocytopathologic confirmation of malignant cells in the pleural fluid or tissue. Pleural fluid cytologic examination is the first line of investigation in suspected cases and has a sensitivity up to 60% (dependent on tumor type, extent of disease, and experience of the cytologist). Immunocytochemical analysis plays an invaluable role in cytologic assessment of pleural fluid, improving the differentiation between benign and malignant effusions and between different malignancies (particularly metastatic adenocarcinoma versus mesothelioma). If malignancy is suspected, generally little incremental benefit is obtained from cytologic analysis of pleural fluid on more than two occasions.

Mesothelioma, the most common primary malignant tumor of the pleura, often is challenging to diagnose. Research in recent years has proposed several potential adjunct diagnostic biomarkers. Mesothelin is a U.S. Food and Drug Administration (FDA)-approved diagnostic and prognostic marker for mesothelioma, although it is not recommended as a sole diagnostic marker (having sensitivity of 48% to 84% and specificity of 70% to 100%). False-negative results may occur with sarcomatoid mesothelioma, which often does not express mesothelin. A raised mesothelin level also can be found with other malignancies (most commonly, pancreatic or ovarian carcinoma but also lung adenocarcinoma and lymphoma). Hence, in patients with elevated pleural fluid or blood mesothelin, further investigations should be considered even if cytologic examination demonstrates no malignant cells. Megakaryocyte-potentiating factor (MPF), which originates from the same precursor protein as for mesothelin, has been shown to have a similar diagnostic performance for detection of mesothelioma. After initial interest in the use of osteopontin for mesothelioma identification, subsequent data have shown it to have a lower diagnostic accuracy than mesothelin.

Other tumor markers and pleural fluid cytokine measurements are neither sensitive nor specific enough for clinical use. Cytogenetic evaluation may be a helpful addition to characterize chromosomal markers of hematolymphoid and mesenchymal malignancies. Flow cytometry of the pleural aspirate should be considered if lymphoma is a possibility.

Gram staining and culture of pleural fluid should be performed in an appropriate clinical setting. Direct inoculation of bottled culture media (“blood culture bottles”) along with separate microbiologic analysis of pleural fluid in a universal container has been shown to increase microbial yield. In cases characterized by presence of frank pus, the diagnosis of empyema is secured, but Gram staining and culture may help to identify the causative organism(s) and to guide therapy.

Testing for tuberculosis should be conducted if clinically indicated, particularly if the fluid is lymphocyte-predominant (more than 50% of total leukocytes). Direct smears and aspirate cultures have low sensitivities (less than 5% and 10% to 20%, respectively) because tuberculous pleuritis usually develops as a type IV hypersensitivity reaction, with low mycobacterial load in the pleural cavity. Demonstration of caseating granulomas by closed or thoracoscopic pleural biopsy clinches the diagnosis. Debate exists regarding the role of adenosine deaminase (ADA), interferon-γ, interferon-γ release assays (IGRAs), and polymerase chain reaction (PCR) techniques in diagnosing tuberculous pleuritis. ADA, an enzyme present in lymphocytes, commonly is measured in disease-endemic countries, and a high pleural fluid ADA level suggests tuberculous effusions, with a sensitivity of greater than 90%. False-positive results have been reported (especially with empyema, rheumatoid effusions, and lymphoma). In nonendemic areas, a low ADA measurement may help rule out tuberculous pleuritis in some patients. Assays for the isoenzyme ADA_2, which predominates in tuberculous effusions, are not widely available. Unstimulated interferon-γ pleural fluid levels perform about as well as ADA for diagnosis in this setting but are more expensive. Peripheral blood and pleural fluid IGRAs have been shown to be of limited clinical value in the diagnosis of tuberculous pleural effusions.

A raised amylase level (above the serum upper limit of normal or if the pleural fluid–to–serum ratio is more than 1) in an appropriate clinical setting can help confirm effusions from esophageal rupture and pancreatic diseases (acute pancreatitis or pseudocyst). Isoenzyme analysis may further differentiate the source of amylase (salivary or pancreatic) but is rarely needed. A raised amylase is present in some malignant (usually adenocarcinomas) effusions.

In suspected cases of chylothorax, lipoprotein electrophoresis for chylomicrons or a triglyceride concentration greater than 1.24 mmol/L (110 mg/dL) confirms the diagnosis. The presence of cholesterol crystals at microscopy with a pleural fluid cholesterol level more than 5.17 mmol/L (200 mg/dL) is diagnostic of a pseudochylothorax. Chylomicrons are not found in pseudochylothorax fluid.

Pleural fluid levels of rheumatoid factor and antinuclear antibody mirror serum values and add little to clinical management of rheumatoid or systemic lupus erythematosus–related pleuritis. Complement levels can be reduced but are of little clinical value. Very low pleural fluid glucose levels typically are seen in patients with a rheumatoid effusion, and differentiation from empyema fluid may be difficult.

β₂-Transferrin is found in cerebrospinal fluid, and its presence in pleural fluid confirms presence of a duropleural fistula. Raised creatinine values are seen in urinothorax.

Percutaneous Biopsy

“Blind” percutaneous pleural biopsy (performed using an Abrams or Cope needle) has been shown to be inferior to imaging-guided biopsy for the diagnosis of malignancy in a randomized controlled trial. Blind biopsy can still be useful in diagnosing diseases with diffuse pleural involvement, especially granulomatous pleuritis (e.g., from TB or rheumatoid arthritis), particularly if thoracoscopic facilities are not readily available. The yield, however, is critically dependent on operator experience and also on the number of passes.

Thoracoscopy

Thoracoscopy is indicated in patients with an undiagnosed exudative pleural effusion after a nondiagnostic thoracentesis. Thoracoscopy can be performed by physicians or surgeons either using local anesthesia and sedation (pleuroscopy) or with the patient under general anesthesia using single-lung ventilation (video-assisted thoracoscopic surgery, [VATS]). Pleuroscopy allows direct visualization of almost the entire pleural surface (Figure 69-9), drainage of fluid, parietal pleural biopsy, and pleurodesis by talc poudrage in the same procedure. Diagnostic sensitivity approaches 95% in malignant pleural diseases and almost 100% in tuberculous pleurisy. Complications are few, and mortality rates low (less than 0.5%). VATS has a similarly high diagnostic sensitivity but also allows other invasive interventions during surgery, including lung biopsy. For details of the technique, refer to Chapter 74.

Figure 69-9 A, Thoracoscopic view into the right pleural cavity showing multiple tumor nodules covering the parietal pleura. B, Thoracoscopy allows identification and biopsy of pleural abnormalities under direct vision while avoiding potentially hazardous structures (e.g., highly vascularized adhesion).

Thoracotomy

Open biopsy with thoracotomy is now seldom indicated. It is a more invasive procedure with significant morbidity, especially postthoracotomy pain.

Thoracentesis

Therapeutic pleural aspiration can be performed at the bedside to relieve breathlessness. For recurrent effusions (e.g., malignant pleural effusions), a definitive procedure to stop fluid reaccumulation, such as pleurodesis, should be considered. If this fails, repeat thoracenteses may be necessary for palliation of breathlessness.

Placement of Indwelling Pleural Catheters

The insertion of a long-term tunneled indwelling pleural catheter (IPC) allows outpatient drainage of refractory malignant pleural effusions. The ambulatory catheters can be inserted as a “day case,” avoiding the need for hospital admission. Its presence induces a complete or partial pleurodesis in up to 70% of patients, and the patient (or the caregiver) can drain the effusion as guided by symptoms, averting hospitalization. Implantation of an IPC also is indicated in patients with a symptomatic effusion and underlying trapped lung, as described later. The catheters usually are well tolerated in clinical practice, but reported complications include development of pleural septations, pleural or soft tissue infection, local tumor invasion at the insertion site, and catheter displacement (Figure 69-10). A randomized trial is under way to evaluate the use of IPCs as first-line therapy in malignant effusions.

Figure 69-10 Magnified chest radiograph demonstrating a left-sided indwelling pleural catheter (IPC), which has become partly displaced and now has a drainage hole within the subcutaneous tissues (arrow).

Insertion of a pleuroperitoneal shunt is another alternative if pleurodesis fails. This intervention is contraindicated in the presence of ascites, and potential complications include infection and shunt occlusion (approximately 10%).

Noninvasive Palliation

In terminally ill patients with symptomatic effusions, use of opiates and oxygen therapy to alleviate breathlessness may be appropriate.

Pleurodesis

The dual aim of pleurodesis is to achieve fusion between the visceral and parietal pleural surfaces and to prevent reaccumulation of pleural fluid (or air). It is indicated in symptomatic malignant pleural effusions and occasionally for recurrent benign effusions when other treatments fail. Dyspnea in patients with malignant effusions often is multifactorial. Only patients with symptomatic improvement consequent to evacuation of the effusion should be considered for pleurodesis.

To achieve successful pleural symphysis, pleural fluid is drained to allow apposition of the pleural layers. Fluid can be drained either at thoracoscopy or using a chest tube. Use of smaller-bore chest tubes (10 to 14F) is associated with less discomfort and allows equal pleurodesis efficacy compared with larger tubes. Pleurodesis, accomplished by mechanical abrasion or using chemical agents, aims to induce damage to the pleura. The resultant acute pleural inflammation, if sufficiently intense, will progress to chronic inflammation with consequent development of pleural adhesions and fibrosis. Pain and fever are common, secondary to the inflammatory process. In animal studies, dampening of this inflammatory process with corticosteroids inhibits pleurodesis. Whether this applies in humans remains unclear.

Pleurodesis is not indicated if close contact of the pleural layers cannot be achieved. The presence of a so-called trapped lung, when lung expansion is restricted either by visceral pleura tumor encasement or by endobronchial obstruction, prohibits effective pleurodesis (Figure 69-11). With “entrapment” of a significant (greater than 50%) proportion of the lung, insertion of an indwelling pleural catheter should be considered.

Figure 69-11 Radiographic appearance of trapped lung. The patient had bronchogenic adenocarcinoma and initially was seen with a massive right-sided pleural effusion causing mediastinal shift (A). After pleural drainage (with an intercostal tube in situ), his lung failed to reexpand (B). Recognizing the presence of trapped lung is important, because pleurodesis not only is unlikely to be successful but may induce visceral fibrosis, causing further restriction of the underlying lung.

Talc is the most commonly used pleurodesis agent worldwide. It can be insufflated as a dry powder (poudrage) during thoracoscopy or instilled through a chest tube as a slurry, with similar efficacy of approximately 75% in a large randomized study of 482 patients. Adverse effects (e.g., pain, fever) are common after talc pleurodesis and can be serious. The incidence of reported talc-induced acute respiratory distress syndrome (ARDS) ranges from 0 to 9%, and its mechanism remains poorly understood. Particle size, presence of contaminants, and reexpansion pulmonary edema have all been implicated. A randomized controlled trial has confirmed that nongraded talc preparations with “small” particle size (less than 15 µm median diameter) led to more systemic and pulmonary inflammation with resultant hypoxemia. A subsequent study using larger-particle graded talc found no evidence of associated ARDS. In patients with preexisting respiratory compromise, alternative agents (e.g., doxycycline or bleomycin) should be considered. Other agents such as povidone-iodine, silver nitrate, and OK432 are used in some centers.

Congestive Cardiac Failure

Effusion secondary to congestive cardiac failure accounts for most transudative effusions worldwide. It is also the most common cause of pleural effusions necessitating acute medical hospital admissions, and as many as 50% to 75% of the hospitalized patients will have pleural effusions, commonly bilateral. Such effusions are thought to arise from transpleural migration of pulmonary interstitial fluid accumulated from elevated pulmonary capillary pressure.

Clinical history and physical examination usually secure the diagnosis. Chest radiography commonly reveals bilateral effusions, cardiomegaly, and venous congestion.

Management is directed at the underlying congestive cardiac failure. Pleurodesis has been performed in refractory cases with variable success. If the patient fails to respond to diuresis or other, atypical features (i.e., unilateral effusion, massive pleural effusion, or fever) are present, further investigation is warranted.

Occasionally, the pleural fluid may be classified as an exudate using Light’s criteria, probably arising because of differential drainage of fluid over protein molecules from the pleural space as a result of diuretic therapy. A plasma-to-effusion protein gradient of more than 3.1 g/dL suggests that the fluid originally formed as a transudate. N-terminal (amino-terminal) pro-brain natriuretic peptide (NT-proBNP) is a sensitive marker of cardiac failure, in both blood and pleural fluid, and can be used to avoid repeated investigations in patients with exudative effusions suspected to be originally caused by cardiac failure. In patients with bilateral transudative pleural effusions with pulmonary hypertension and normal left ventricular function, pulmonary venoocclusive disease (PVOD) should be considered.

Hepatic Hydrothorax

Hepatic hydrothorax occurs in approximately 5% of patients with liver cirrhosis and portal hypertension. Pleural effusions accumulate from transdiaphragmatic migration of ascitic fluid. Fluid also accumulates from the decreased oncotic pressure associated with secondary hypoalbuminemia, and extremely rarely from hemorrhage of congested collateral veins. Most hepatic hydrothoraces are right-sided, and concomitant ascites is frequent but is not a prerequisite. The diagnosis is based on the presence of cirrhotic liver disease, ascites, and transudative fluid on thoracentesis. Spontaneous bacterial empyema may occur and, even with prompt recognition, carries a mortality rate ranging up to 20%.

Management should be directed at treating portal hypertension and reducing the volume of ascitic fluid with diuretic therapy and sodium restriction. This approach frequently fails, and therapeutic thoracenteses may be indicated for symptomatic relief. Loss of electrolytes and protein is a concern. Pleurodesis may be tried in therapy-resistant cases, and repair of any diaphragmatic defects attempted at thoracoscopy. Use of concomitant continuous positive airway pressure (CPAP) ventilation at the time of pleurodesis to reverse the peritoneal-pleural gradient has been successful in case reports but remains experimental at present. Transjugular intrahepatic portosystemic shunting (TIPS) can relieve symptomatic hepatic hydrothorax, but liver transplantation, in selected patients, is the only definitive therapy.

Renal Effusions

Pleural effusions may arise in association with nephrotic syndrome, acute glomerulonephritis, or peritoneal dialysis. Transdiaphragmatic migration of peritoneal dialysate often results in a right-sided effusion chemically resembling the dialysis fluid.

Urinothorax is rare but should be considered whenever a low-pH transudative effusion is encountered in clinical practice. It arises secondary to renal obstruction with resultant retroperitoneal urine collection and transfer into the ipsilateral pleural cavity, or a fistula draining into the pleural cavity. An ammonia-like smell is diagnostic, and increased fluid-to-serum creatinine levels are confirmatory.

Parapneumonic Effusion and Empyema

Pleural effusions occur in up to 57% of patients with pneumonia and range in size from small subcentimeter effusions to large effusions causing ventilatory embarrassment. A majority of these (approximately 60%) are sterile “simple” parapneumonic effusions reflecting vascular hyperpermeability from pleural inflammation secondary to the underlying pneumonia. “Complicated” parapneumonic effusions are characterized by neutrophilia, low pH and glucose, and often fibrinous septations (caused by fibrin deposition within the procoagulant milieu of the infected pleural space). The current belief is that complicated pleural effusion represents one end of the spectrum of pleural infection and should be treated as such. At the other end of the spectrum is empyema, characterized by frank pleural pus or presence of bacteria in the pleural fluid (Table 69-3).

Differentiation between simple and complicated effusions is important, because an infected pleural space requires prompt pleural drainage, being associated with significant morbidity and mortality (15% of patients require surgery, and up to 20% die). Parapneumonic effusion should be suspected in all patients with pneumonia, particularly those in whom clinical improvement fails to occur or laboratory markers of infection remain persistently abnormal. Patients with diabetes mellitus, alcohol or substance abuse, coexisting chronic lung disease, immunosuppression, or rheumatoid arthritis have a higher incidence of pleural infection. Poor dental hygiene is more prevalent in those with anaerobic infection. A low serum albumin less than 30 g/L, C-reactive protein more than 100 mg/L, platelet count more than 400 × 10⁹/L, serum sodium less than 130 mmol/L, intravenous drug use, and alcohol misuse have been shown to be associated with development of complicated parapneumonic effusion in patients with pneumonia. Recent novel genetic studies also suggest that a variant of the protein tyrosine phosphatase (PTPN22 Trp620) is associated with susceptibility to invasive pneumococcal disease and gram-positive empyema. Primary empyema, in the absence of pneumonia, is responsible for 4% of pleural infection cases.

Chest radiography usually identifies the pleural collection and concomitant pneumonia. In a patient with sepsis, the finding of a new encapsulated effusion in a nonbasal position is strongly suggestive of pleural infection. Ultrasound imaging is of value in the evaluation of pleural infection, allowing fluid characterization and localization (which is particularly important for loculated fluid). In a patient with signs of infection, septations on ultrasound images suggest an infected pleural space with low pH, low glucose, and high lactate dehydrogenase (LDH); heavily echogenic fluid usually represents pus (or blood). Of note, however, the converse does not hold; absence of such sonographic features does not rule out pleural infection. Pleural-phase CT provides detailed information and discriminates between empyema and lung abscess. Empyemas usually are lenticular in shape, with compression of surrounding lung parenchyma. The “split pleura” sign, with enhanced parietal and visceral pleural tissue visible as the surfaces are separated by the pleural collection, is characteristic (Figure 69-12). Pleural thickening is seen in 86% to 100% of empyemas and in 56% of parapneumonic effusions. Pleural enhancement and increased attenuation of the extrapleural subcostal fat are suggestive of an infected pleural space.

Figure 69-12 A contrast-enhanced computed tomography image of pleural empyema. The “split pleura” sign is demonstrated, with enhancement of both the visceral and the parietal pleural surfaces evident.

(Courtesy Dr. Rachel Benamore, Oxford, United Kingdom.)

Frankly purulent or malodorous fluid confirms presence of an empyema. If not purulent on gross examination, the pleural fluid pH should be assessed. A pH less than 7.2, or reduced fluid glucose, indicates a complicated parapneumonic effusion. Positive Gram staining or fluid culture confirms pleural infection and can guide antibiotic choice but is positive in only about 50% to 60% of patients. The use of genetic microbiologic testing to allow organism identification (using PCR and sequencing of the bacterial 16S ribosomal RNA gene) is under investigation. Treatment of complicated parapneumonic effusions or empyema includes appropriate antimicrobial therapy, optimal nutritional support, venothromboembolism prophylaxis, and evacuation of the infected pleural fluid (by chest tube or surgery).

Bacteriology

It is important to differentiate between community- and hospital-acquired pleural infection, because the bacteriology and prognosis differ for the two types. Hospital-acquired cases often are iatrogenic or secondary to hospital-acquired pneumonia and carry a mortality rate several times higher than that for community-acquired cases. The microbiology also varies geographically. In the United Kingdom, gram-positive streptococci account for most community-acquired cases of pleural infection, with the Streptococcus milleri group (Streptococcus constellatus, Streptococcus intermedius, Streptococcus anginosus) accounting for 24% and Streptococcus pneumoniae causing 21%. Staphylococcus aureus is isolated in 10% of the cases. Gram-negative organisms, such as Haemophilus influenzae, Escherichia coli, Pseudomonas spp., and Klebsiella spp. occur in about 10% of cases, particularly in patients with comorbid conditions such as diabetes. Anaerobic pathogens are detected in 12% to 34% of cases, although some studies suggest a much higher incidence. With hospital-acquired infection, methicillin-resistant Staphylococcus aureus (MRSA) is responsible for more than a quarter of cases of pleural infection, and gram-negative organisms such as E. coli, Enterobacter spp., and Pseudomonas spp. are common and particularly associated with critical illness necessitating admission to an intensive care unit.

Antibiotics

All patients should receive antibiotic therapy. Empirical treatment should be initiated while awaiting bacterial culture results and must take into account where the infection was acquired (hospital or community), patient comorbid conditions, and local bacterial resistance patterns. A common antibiotic choice to cover community-acquired pathogens is a beta-lactam antibiotic in conjunction with a beta-lactamase inhibitor (such as amoxicillin and clavulanic acid) or piperacillin-tazobactam. Metronidazole usually is given as well, to ensure adequate anaerobic coverage. Given that hospital-acquired pleural infection is more likely to be caused by resistant pathogens, including gram-negative enteric bacteria and MRSA, a reasonable antibiotic selection in such patients is a carbapenem with vancomycin. Therapy may be modified in the light of positive culture and sensitivity results. Prolonged courses of antibiotics (e.g., 4 to 6 weeks) often are required.

Drainage

Simple parapneumonic effusions generally do not require drainage, but drainage of complicated effusions and empyema is key to their management.

Loculation is common in pleural infection, and imaging-guided insertion of chest drains can ensure optimal placement. Selection of appropriate chest tube size remains contentious, with no randomized trial evidence available, although recent studies have suggested that smaller (less than 15F) Seldinger inserted tubes are less painful, are associated with fewer complications, and perform as well as larger, blunt dissection–inserted tubes. Instillation of intrapleural fibrinolytic agents for complex parapneumonic effusions or empyema has been proposed to disrupt fibrinous septations and potentially improve drainage. However, a large multicenter study, the first Multicenter Intrapleural Sepsis Trial (MIST1), with 454 subjects, failed to demonstrate a benefit on mortality statistics, need for surgery, or duration of hospital stay for patients who received intrapleural streptokinase. A 2008 Cochrane review examining seven fibrinolytic studies concluded that use of fibrinolytics conferred no significant benefit in reducing surgical requirements. The recent MIST2 study examined combination fibrinolytic-plus-DNase therapy in a 2 × 2 factorial randomized controlled trial of tissue-type plasminogen activator (tPA) and deoxyribonuclease (DNase) in 210 patients. This study found that combination tPA and DNase significantly improved pleural fluid drainage (primary outcome) and led to a reduced rate of surgical referral and shorter hospital stay (secondary outcomes), although further studies are needed to define the size of treatment effects.

Prompt referral for surgical drainage is indicated in the 30% of patients who have ongoing sepsis with chronic pleural fluid effusion despite optimal medical treatment. Secondary pleural thickening (fibrothorax) is seen in less than 10%; in this setting, however, residual pleural thickening and lung function impairment always lessen with time and are not indications for surgery. VATS is preferred, but more invasive procedures (e.g., thoracotomy or rib resection with open drainage) are sometimes needed.

Use of VATS as first-line treatment of pleural infection remains contentious, with randomized studies in adults and children failing to show a significant difference in mortality or major morbidity with VATS. Adult studies have shown a shorter hospital stay with VATS (approximately 11 versus 8 days), but a pediatric study showed no difference in length of stay (primary outcome) but a higher cost of VATS.

Rheumatoid Arthritis

Pleural effusions are the most common thoracic manifestation of rheumatoid arthritis and occur in up to 5% of patients, more frequently men (70%). Often the effusions develop several years after the original diagnosis and in 30% of the cases arise concomitantly with interstitial lung disease or rheumatoid lung nodules. The effusion may persist for years, although spontaneous resorption occasionally is observed. Symptomatic recurrent effusions are rare and may respond to systemic corticosteroid therapy.

Distinguishing rheumatoid effusions from empyema or pseudochylothorax is important but can be difficult. Pleural fluid from a rheumatoid effusion is characterized by leukocytosis, a low glucose, low pH, and increased LDH—features indistinguishable from those of pleural infection. Low complement and elevated pleural fluid rheumatoid factor levels are not diagnostic.

Systemic Lupus Erythematosus

Fifty percent of patients with systemic lupus erythematosus (SLE) will have a pleural effusion, often bilateral and typically exudative (occasionally hemorrhagic), with a lymphocytosis (if the effusion is chronic). Nonsteroidal antiinflammatory medication may relieve pleuritic chest pain, and corticosteroid treatment results in a swift clinical response if the patient remains symptomatic. Pleural effusions arising as a consequence of drug-induced lupus erythematosus are rare.

Other Connective Tissue Disorders

Granulomatosis with polyangiitis (Wegener’s) is associated with pleural effusion in approximately 30% of patients and responds to treatment of the underlying condition with immunosuppressant therapy. Pleural effusion may occur in up to 30% of those with Churg-Strauss syndrome; typically, the fluid is rich in eosinophils. The effusions usually are not of any clinical consequence, and most resolve with corticosteroid therapy.

Hemothorax

Hemothorax is characterized by pleural fluid hematocrit values greater than 50% of peripheral blood. The main causes are trauma, iatrogenic (most commonly after thoracic surgery) factors, malignancy, anticoagulation therapy, pulmonary infarction, and aortic rupture. Complications include empyema (1% to 4% of traumatic hemothoraces), trapped lung, and fibrothorax. Treatment is guided by the clinical scenario, but tube thoracostomy to enable evacuation of the blood and assessment of the extent of hemorrhage is advised. Surgical intervention (preferably VATS) is indicated if the bleeding rate is greater than 200 mL/hour. Prophylactic antibiotics usually are administered, although literature evidence to support this practice is scarce.

Chylothorax

Chylothorax develops when disruption of the thoracic duct results in passage of chyle (lymph rich in chylomicrons) into the pleural cavity. The pleural fluid characteristically is milky in appearance (Figure 69-13), except in starved patients. Chylothorax must be differentiated from pseudochylothorax and empyema fluid (see under “Pleural Fluid Analysis”).

Figure 69-13 Typical appearance of chylothorax drainage.

Trauma is the most common cause of chylothorax, which is predominantly iatrogenic secondary to thoracic duct damage during cardiothoracic surgery (e.g., esophagectomy). An important nontraumatic cause is malignancy, especially lymphoma. Congenital chylothorax arises from thoracic duct malformation, and repair of congenital diaphragmatic defects may cause chylothoraces. Lymphangioleiomyomatosis and nonsurgical trauma rarely produce chylothoraces. Chylothorax is idiopathic in 15% of the reported cases. CT scanning should be performed to exclude lymphoma (and other potential causes) in patients without a history of trauma.

Management of chylothorax aims to optimize nutrition, relieve symptoms, and close the thoracic duct defect. Malnourishment secondary to chyle loss can be disabling. Total parenteral nutrition, or a low-fat diet with medium-chain fatty acids, should be adopted to reduce chyle flow. If the patient is symptomatic with dyspnea, fluid can be removed by thoracentesis or chest tube drainage. Pleuroperitoneal shunts have been recommended in the absence of coexisting chyloascites. Pleurodesis has been attempted in refractory cases. With traumatic chylothorax, the defect frequently closes spontaneously. Use of octreotide, a somatostatin analogue, also has been reported, mainly in pediatric-based case series, to enhance thoracic duct closure. Percutaneous thoracic duct embolization performed under radiologic guidance can be attempted before consideration of surgical thoracic duct ligation by VATS or thoracotomy. This approach initially was described in 1998 and offers a minimally invasive treatment strategy for those patients who are unable to tolerate surgical intervention. Catheterization of the thoracic duct is a prerequisite for its use, and anatomic obstruction of the cisterna chyli or retroperitoneal lymphatic chain obviates its use. In patients with underlying lymphoma, radiotherapy or systemic chemotherapy may control the chylothorax.

Pseudochylothorax

Pseudochylothorax, or cholesterol pleurisy, is most commonly associated with tuberculous pleurisy and rheumatoid pleuritis. The pleural fluid has a very high cholesterol content and no chylomicrons. Demonstration of cholesterol crystals is diagnostic. Pseudochylothorax was conventionally thought to result from the accumulation of cholesterol and lecithin-globulin complexes within long-standing pleural effusions enveloped by thickened fibrotic pleura. Recent evidence, however, suggests that such chronicity and pleural fibrosis are not prerequisite in pseudochylothorax from rheumatologic causes. Specific treatment for the condition is seldom indicated. Dyspnea is more likely to be a result of the pleural fibrosis and is seldom relieved by thoracentesis. Respiratory insufficiency, infection, and bronchopleural or pleurocutaneous fistulas are rare complications.

Cystic Fibrosis

Development of pneumothorax in patients with cystic fibrosis (CF) carries an important prognostic implication; median survival after pneumothorax is 30 months. The pneumothorax itself is not the cause of demise but is a surrogate marker of end-stage lung disease, because patients with a forced expiratory volume in 1 second (FEV₁) of less than 30% predicted are most at risk. Occurrence of pneumothorax usually is unrelated to acute respiratory illness. Adherence to the standard guidelines for management is recommended, however, with concomitant antibiotic use to avoid secondary infection. Pleurodesis to prevent recurrence of pneumothorax is important, because recurrence rates are approximately 50%. Early liaison with the transplantation team is appropriate if surgical input is required.

Pulmonary Langerhans Cell Histiocytosis

Spontaneous pneumothorax may be the presenting feature in up to 25% of patients with Langerhans cell histiocytosis (LCH). Treatment options are those discussed previously for pneumothorax, together with management of the underlying LCH (see Chapter 54).

Lymphangioleiomyomatosis

In addition to chylothorax, patients with lymphangioleiomyomatosis (LAM) are at risk for development of pneumothorax. This complication is common in this population (affecting up to 80%) and may be the presenting feature of the underlying LAM. Recurrent pneumothoraces occur in two thirds of patients, and early definitive intervention is recommended to reduce recurrence rates. Pleurodesis may increase the rate of perioperative bleeding (occurring as a result of resultant adhesions) but does not preclude successful future lung transplantation (see Chapter 54).